Read multiple .CSV files with csvread using a for loop

Hello all, I'm have some trouble developing an algorithm to read multiple files at once and store them in different matrices. Heres a little, but sufficient, description of the problem:

Say I have n files named i_p.CSV, where n is a large enough number to do this procedure manually and i is a number that goes from 1 to n. The file i_p.CSV is a csv file with three columns, so what I would end up have on MATLAB is a matrix called data_i, which would be a (m x 3) matrix, where m is also a very large number.

Any tips on how to proceed?

Answer 1

[dpb]

d=dir('*_p.clv');   % dir struct of all pertinent .csv files
n=length(d);        % how many there were
data=cell(1,n);     % preallocate a cell array to hold results
for i=1:n
  data(i)=csvread(d(i).name);  % read each file
end

Now, iff every one of the n .csv files has the same number of rows you can convert to an array...

data=cell2mat(data);

If they aren't all the same length you'll have to augment the shorter to the length of the longest or truncate the longer to that of the shortest.

[Jorge Mira Pérez]

dpb, it worked mate! Thanks a lot! Sorry for taking this long to answer I've been pretty busy. In case someone else tried this, for some reason I recommend you guys to add path before! Something like

addpath('/...')

[Spresi]

Hey there, I am having the same issue and I am not able to read all of the csv files.

The part of code that I dont get is:

data(i)=csvread(d(i).name);

what would "name" be in this case?

Thank you in advance..

[dpb]

name is the filename field in the directory information structure returned by dir().

Each iteration thru the loop will return the next filename found and returned in struct d

[Spresi]

I am still unable to read all my csv files. So here's my problem:

I have around 9000 folders named ACCEDE00000 to ACCEDE08999 and each of them contains multiple csv files, but I'm only interested in one specific csv file named frame_table1_final.csv for each of 9000 ACCEDE videos. So I tried:

d=dir('/ACCEDE*/frame_table1_final.csv');   
n=length(d);        
data=cell(1,n);     
for i=1:n
  data(i)=csvread(d(i).name);  
end

The error I get:

Error using csvread (line 35)
File not found.

Error in emotionn (line 5)
    data(i)=csvread(d(i).name);

P.S. I am willing to import them as numeric arrays. Each of the csv files has same number of columns but different number of rows.

Thank you again!

[dpb]

Per the doc for dir() the filename is in .name; the folder/subdirectory name is in the .folder field.

If the file(s) are not in the working directory or on the matlabpath, you must fully-qualify them...

 data(i)=csvread(fullfile(d(i).folder,d(i).name));  

[Yuvarajendra A Reddy]

Hey all.. I completely understand reading Multiple .csv files. But i have a different issue. When I'm reading these files one by one, I need to skip certain lines in the begininning ( whicjh i managed to do so by using 'HeaderLines' command) and a last line, specifically the last row. Could someone help me sort this out please..

[dpb]

No truly practical way to skip the last line; simply delete it from the read array after reading the file.

[Gershon Koral]

This answer is incorrect. The for loop should rather read:

data{i} = csvread(d(i).name);